Tip:
Highlight text to annotate it
X
Create a new project
Define factors
Typhical factors are Quantitative and Controlled
Click or use tab to browse through the window
Define responses
Min, Target and Max values are used in the optimizer
Transformation on some responses can improve the model
Select Response Surface Modelling (RSM) as objective
Choose the recommended design
A worksheet is now created. Perform experiments in the order defined by "Run Order".
Insert the results
It is also possible to copy/paste in results from MS Excel, for instance
Fit the model
Evaluate and improve the model per response
Begin with raw data inspection
Model summary statistics as a graph
Remove non significant terms to improve the model in the interactive coefficient plot
Notice the increase of the Q2 value (predictive R2)
Residual inspection. Stright line = white noise
Observed versus predicted
Continue with the next response
Now use the models.
Create a contour plot.
Results in 2D and 3D
Use the optimizer to find the optimal running condition
The resulting optimal running condition is...
Air = 261 kg/h, EGR=9.2 % and Needlelift = -3.16 BTDC
The optimal setting can be displayed in a sweet spot plot
Green area = All specifications fulfilled